Excessive optimization will run in actual combat

Middle of the night, there are a few people at work, have entered the dream of incense, and in the last article I introduced to the Martin trading strategy, before the actual combat is to run the market, rearranging ideas, calculated 5 months of data, this Late to write more than 300 lines of code, back to test a few months of data, there are more than 30 pit did not fill, and the movement stop function is not added, mobile stop to join, the capital curve should increase 5-10 times, The current test only caught 50 points, should be very perfect.

USDJPY Monthly data retest

Ran out of the map although very comfortable, but in actual combat, there will be different pit out, resulting in loss or less earn, the transaction in the K line is changing at any time, no time is the same trend, the Martin trading strategy I do not It is recommended that you go by hand, the probability will run the market, improved, the probability of success increased significantly, the number of failures reduced in the actual deployment to the server before the need for a lot of testing


Just in the test time, encountered this kind of graphics, which is often said to play face graphics, foreign exchange futures are often rising and falling after a platform for finishing, and then decide the direction, before the start will often ventilate, down the trap Face up, and then pull up, or lure more down to play, which can do more, but also short the trend to play out, is to do or do not often risk greater than the opportunity, if not in the process to deal with here Destined to lose money, this treatment is not optimized, it is necessary to fill the pit, many of us can not think of the pit, someone on the network to sell their own strategic procedures, very beautiful financial curve, but in fact these applications run but the market , That is because over-optimization, writing procedures need to be optimized, but do not need to be over-optimized.

In the absence of mobile stop to run out of this result, is very good, the success is greater than the number of failures, retracement small, tomorrow to continue this strategy patch.